A New Constraint on the Escape Fraction in Distant Galaxies Using Gamma-ray Burst Afterglow Spectroscopy

TL;DR

This paper introduces a novel method using gamma-ray burst afterglow spectra to measure the escape fraction of ionizing radiation from distant galaxies, providing new constraints especially for low-mass galaxies at high redshift.

Contribution

The study presents a new technique that avoids background subtraction issues and prior spectral knowledge, enabling the first constraints on escape fractions in low-mass, high-redshift galaxies.

Findings

Estimated mean escape fraction <fesc>=0.02±0.02

Placed a 95% confidence upper limit <fesc> ≤ 0.075

Sample of 27 GRBs at z>2 used for statistical analysis

Abstract

We describe a new method to measure the escape fraction fesc of ionizing radiation from distant star-forming galaxies using the afterglow spectra of long-duration gamma-ray bursts (GRBs). Optical spectra of GRB afterglows allow us to evaluate the optical depth of the host ISM, according to the neutral hydrogen column density N(HI) observed along the sightlines toward the star-forming regions where the GRBs are found. Different from previous effort in searching for faint, transmitted Lyman continuum photons, our method is not subject to background subtraction uncertainties and does not require prior knowledge of either the spectral shape of the host galaxy population or the IGM Lya forest absorption along these GRB sightlines. Because most GRBs occur in sub-L_* galaxies, our study also offers the first constraint on fesc for distant low-mass galaxies that dominate the cosmic luminosity density. We have compiled a sample of 27 GRBs at redshift z>2 for which the underlying N(HI) in the host ISM are known. These GRBs together offer a statistical sampling of the integrated optical depth to ionizing photons along random sightlines from star-forming regions in the host galaxies, and allow us to estimate the mean escape fraction <fesc> averaged over different viewing angles. We find <fesc>=0.02\pm 0.02 and place a 95% c.l. upper limit <fesc> <= 0.075 for these hosts. We discuss possible biases of our approach and implications of the result. Finally, we propose to extend this technique for measuring <fesc> at z~0.2 using spectra of core-collapse supernovae.